This market resolved against "No" on 2026-06-29, verified on-chain against Polymarket's settlement — the polymax engine called it wrong, and logs the loss anyway.
On 2026-06-27, the polymax engine's sprint screen flagged this market and took "No" at 75¢ — the market was pricing that outcome at 75%, so a correct call would pay 1.34× the stake. 1.9 days later the market resolved on-chain, and the position settled as a loss: −$20.42, the full stake.
sprint is one of the engine's ground-truth strategies: across 97 out-of-sample historical positions it hit 78.4% (95% CI 69.2–85.5%), a +3.8-point edge over the market's implied odds, returning +5.1% per bet. This position was sized with fractional Kelly from those numbers — not from a hunch.
The engine publishes its losses on purpose: a 78.4%-hit-rate strategy still loses 22% of the time, and any track record without losses on display is marketing, not measurement. This one is permanently archived on the losses page.
Paper-trading position settled against the gamma-api resolved outcome. This page was minted automatically when the market resolved and regenerates from the engine's ledger. Past performance does not guarantee future results. Not financial advice.